From Bread, Cheese, and Kimchi to Pyruvate and NAD⁺
by Professor Stephen T. Abedon Steve Abedon, Ph.D., has taught well over 100 individual classes and ~10 individual courses (including introductory biology, microbiology, physiology, biochemistry, etc.) over his 30+ years as a college professor. He has over 150 publications, including multiple monographs, and is ranked in the top 1 or 2 percent among researchers of all disciplines worldwide. His research focuses on bacteriophage biology and phage therapy. (abedon.1@osu.edu)
phage.org | phage-therapy.org | biologyaspoetry.org | abedon.phage.org | google scholar | publications
Jump to: 🧀 Fermented Foods | 🍞 Dough Balls | 🧪 What Fermentation Is | 🤔 Why Ferment? | 🔀 Types of Fermentation | ⚗️ Fermentation Products | 🧮 More Calculators
biologyaspoetry.org/laboratories/fermentation/ · Abedon’s Books
How can I improve this page? contact: fermentation@biologyaspoetry.org
People have been fermenting food for thousands of years, long before anyone knew that microbes existed. The appeal is partly flavor and partly safety: many fermentation products, especially acids and alcohol, make food inhospitable to the organisms that would otherwise spoil it or make us sick. A selection of them is worth tasting, if you are willing (the tasting is optional!), and the videos below introduce them. The first three are mine; the others are by other people and approach fermented foods from other angles.
Each food below lists the kinds of organisms responsible, the fermentation pathways involved, and the main fermentation products. Pathways and products are links, to the Types of Fermentation and Fermentation Products tabs, and those tabs link back here. Use the filters to ask questions such as which foods depend on molds or which foods contain propionic acid. Foods with videos have a thumbnail; click it to play the video within the card.
Baker’s yeast, Saccharomyces cerevisiae, ferments the sugars in dough: a small amount present in flour, more (mostly maltose) released from starch by the flour’s own enzymes, and whatever sugar the baker adds. The CO2 is trapped by the stretchy gluten network and raises the dough, and nearly all of the ethanol evaporates during baking. The Dough Balls tab is this food as an experiment.
A sourdough starter is a stable community of wild yeasts and lactic acid bacteria, with the bacteria typically far outnumbering the yeasts. The yeasts supply most of the CO2, while heterolactic bacteria such as Fructilactobacillus sanfranciscensis (formerly Lactobacillus sanfranciscensis, named for San Francisco sourdough) make the lactic and acetic acids that give the bread its sourness.
Yeast cannot ferment starch directly, so grain is first malted: sprouting activates the grain’s own enzymes, which later break its starch into fermentable sugars, mainly maltose. Ales are typically made with Saccharomyces cerevisiae and lagers with S. pastorianus. Some styles, such as Belgian lambics and other sour beers, also rely on lactic acid bacteria and wild yeasts.
Yeasts ferment the sugars of grape juice to ethanol and CO2; in still wines the CO2 is allowed to escape, and in sparkling wines some of it is kept. Many red and some white wines then undergo malolactic fermentation by Oenococcus oeni, which converts tart malic acid into softer lactic acid.
Whiskey, rum, vodka, and other spirits begin as fermented liquids, in effect beers or wines, whose ethanol is then concentrated by distillation. Distillation is a physical separation that takes advantage of ethanol’s lower boiling point; it is not itself a fermentation.
Sweetened tea fermented by a community of yeasts and acetic acid bacteria (often with some lactic acid bacteria). The yeasts make ethanol and CO2; the acetic acid bacteria oxidize ethanol to acetic acid and glucose to gluconic acid, and one of them, Komagataeibacter, spins the rubbery cellulose mat floating on top.
Included here as a warning about words. Tea makers call the step that turns green leaves black “fermentation,” but it involves no microbes: crushing the leaves lets their own enzyme, polyphenol oxidase, oxidize the leaves’ polyphenols using oxygen from the air. Some teas, such as Chinese pu-erh, are genuinely fermented by microbes after processing.
Milk fermented with kefir “grains,” which are not grains at all but rubbery clumps of polysaccharide and protein in which lactic acid bacteria, yeasts, and acetic acid bacteria live together. The result is sour, slightly fizzy, and very slightly alcoholic. Water kefir is a separate, sugar-water fermentation with its own grains.
Made by two lactic acid bacteria, Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus, growing together in warm milk. They ferment lactose to lactic acid, and the falling pH makes the milk protein casein coagulate into a gel. Acetaldehyde made by the bacteria is a large part of yogurt’s aroma.
Cream (for sour cream) or milk (for cultured buttermilk) fermented by lactic acid bacteria such as Lactococcus lactis, often together with Leuconostoc. Besides lactic acid, some of these bacteria make diacetyl from the citrate in milk, which is the source of the buttery flavor.
Nearly all cheeses start with lactic acid bacteria (the starter) acidifying milk, usually along with the enzyme rennet, which curdles it. What happens during ripening depends on the cheese: Penicillium roqueforti makes the blue veins of blue cheeses, P. camemberti the white rind of Camembert and Brie, bacteria such as Brevibacterium linens the orange rind of smear-ripened cheeses, and in Swiss-type cheeses Propionibacterium freudenreichii ferments leftover lactic acid, making the CO2 that forms the holes.
Shredded cabbage is mixed with salt, which draws water out of the cabbage cells by osmosis to make a brine. Fermentation then proceeds in succession: heterolactic Leuconostoc mesenteroides first, making CO2 and acetic acid as well as lactic acid, followed by more acid-tolerant homolactic species such as Lactiplantibacillus plantarum. Much supermarket sauerkraut is pasteurized, and some is simply made with vinegar rather than fermented.
Salted napa cabbage or radish seasoned with chili, garlic, ginger, and, in many recipes, fermented seafood. As with sauerkraut, heterolactic bacteria such as Leuconostoc and Weissella dominate early, which is why young kimchi is fizzy; kimchi fermented at cold temperatures stays in that stage longer.
Cucumbers left in salt brine, with dill and garlic, for only a few days, so that they stay crisp and partly green; full sours ferment for weeks. The sourness comes from lactic acid bacteria. Most shelf-stable supermarket dill pickles are instead packed in vinegar, which makes them pickled but not fermented.
Fresh olives are inedibly bitter, because of the compound oleuropein. Curing removes or breaks down the bitterness, whether by soaking in lye (as for Spanish-style green olives), by long soaking in water or brine, or by dry salting. In brine-cured olives, lactic acid bacteria and yeasts ferment the sugars that leach out of the fruit.
Ground meat is mixed with salt, a little sugar, and usually nitrite, then held warm so that lactic acid bacteria (often Pediococcus or Latilactobacillus sakei, formerly Lactobacillus sakei) can acidify it. Acid, salt, and later drying together preserve the sausage. The nitrite is there mainly to prevent botulism: it keeps spores of Clostridium botulinum, which would otherwise find the airless interior of a sausage congenial, from growing and making their toxin. (Botulism is in fact named for sausages, from the Latin botulus, after outbreaks traced to sausages in 18th- and 19th-century Germany.) Acid alone reliably stops C. botulinum only below about pH 4.6, and fermented sausages are usually somewhat less acidic than that, which is why salt, nitrite, acid, and drying are used together rather than any one of them alone. Other bacteria, such as Staphylococcus species, contribute color and flavor, and some styles are coated with a white surface mold.
Raw pork, pork skin, cooked sticky rice, garlic, chilies, and salt are packed tightly and wrapped, traditionally in banana leaves, and left for a few days. The rice and garlic supply sugars that lactic acid bacteria ferment, making the sausage distinctly sour.
Made in two stages. First the mold Aspergillus oryzae (or A. sojae) is grown on soybeans and wheat to make koji, whose enzymes break proteins into amino acids and starch into sugars; this mold stage is aerobic. The koji is then mixed with strong salt brine into a mash (moromi) in which the salt-tolerant lactic acid bacterium Tetragenococcus halophilus and the yeast Zygosaccharomyces rouxii ferment for months. Tamari, made with little or no wheat, is said to have begun as the liquid that collects in fermenting miso.
A paste of soybeans, salt, and koji (the mold Aspergillus oryzae grown on rice, barley, or soybeans), aged from weeks to years. As with soy sauce, the koji’s enzymes do much of the work, with salt-tolerant lactic acid bacteria and yeasts contributing flavor.
Cooked soybeans bound into a firm cake by the white threads of the mold Rhizopus, originally from Java. The mold grows aerobically, digesting proteins and fats, so tempeh is a fermented food by the food definition but involves little fermentation in the biochemical sense (bacteria can acidify the beans somewhat while they soak).
Japanese soybeans fermented by Bacillus subtilis (var. natto), which grows aerobically on the beans and produces the strings of sticky polyglutamic acid that natto is known for.
Vinegar takes two steps: yeasts ferment sugars to ethanol, and then acetic acid bacteria oxidize the ethanol to acetic acid. That second step uses oxygen and an electron transport chain, so despite its usual name it is not a fermentation in the biochemical sense. The word vinegar comes from the French vin aigre, “sour wine.”
Chilies are ground with salt into a mash that lactic acid bacteria ferment; for Tabasco sauce the mash is aged in oak barrels for up to three years before vinegar is added. Fermentation adds sourness and depth of flavor beyond what vinegar alone would give.
Fruits or vegetables, such as mango, lime, or chili, preserved with salt, spices, and often oil, and traditionally cured in the sun. Some traditional pickles undergo lactic acid fermentation during curing; many others are preserved mainly by salt, oil, and added acid, so not every Indian pickle is fermented.
Anglo-Indian chutneys, such as mango chutney, are typically cooked with vinegar and sugar. The fermentation products in them therefore arrive mostly by way of the vinegar, which was itself made by yeasts and acetic acid bacteria, rather than from fermentation in the jar.
Garum was ancient Rome’s favorite condiment: fish, often including their innards, layered with large amounts of salt and left for months, often in the sun. The salt keeps most microbes from growing, while digestive enzymes from the fish’s own guts and flesh break its proteins down into amino acids, glutamate among them, leaving an intensely savory, amber liquid. Salt-tolerant microbes that do grow, such as the lactic acid bacterium Tetragenococcus halophilus (which also works in soy sauce), add acids, aroma, and some protein-digesting enzymes of their own. Southeast Asian fish sauces, such as Vietnamese nước mắm, Thai nam pla, and Filipino patis, are made the same way today (patis often as a byproduct of bagoong, a paste of salted, fermented fish or shrimp), and Italy’s colatura di alici, made from anchovies, is a direct descendant.
First sold in the 1830s by John Lea and William Perrins, chemists in Worcester, England, reputedly from a recipe that a local nobleman brought back from his travels and that tasted awful until it had sat in their cellar for a couple of years. Today the onions and garlic are pickled in malt vinegar for months, and anchovies are cured in salt for about two years, breaking down much as the fish in garum do, mainly by their own enzymes with help from salt-tolerant bacteria. These are then combined with more vinegar, molasses, tamarind, and spices and matured further before the sauce is pasteurized.
Cacao beans are heaped or boxed together with the sugary pulp that surrounds them in the pod, and over several days a succession of microbes ferments that pulp: yeasts first, making ethanol, then lactic acid bacteria, and then acetic acid bacteria, which oxidize the ethanol to acetic acid. The acid and heat soak into the beans and kill the seed embryos, setting off chemical changes inside the beans that produce the precursors of chocolate flavor, which roasting then develops.
Bacterial names follow the 2020 reclassification of the genus Lactobacillus into 25 genera (Zheng et al., 2020), with former names given where they may be more familiar. Sour cream shares its entry with cultured buttermilk.
Your dough-ball predictions and results (from the Dough Balls tab) are kept as you go, and are filled back in if you leave and come back on the same device, though a private window, a different device, or cleared browser data will start you fresh. When you are done, download a summary to keep. It lists your predicted and observed ranks, any measurements, and how your predictions compared with what happened.
Bread is the fermented food you can most easily watch happen. In this experiment you make four very small balls of dough that differ only in whether they contain yeast and whether they contain added sugar, keep them warm for about two hours, and compare how much each one rises. Because they need the time, start the dough balls first and come back to them at the end.
| Treatment | Flour | Yeast water | Plain water | Sugar |
|---|---|---|---|---|
| 1 | 2.5 tsp | — | 1.0 tsp | none |
| 2 | 2.5 tsp | — | 1.0 tsp | a small spatula’s worth |
| 3 | 2.5 tsp | 1.0 tsp | — | none |
| 4 | 2.5 tsp | 1.0 tsp | — | a small spatula’s worth |
Measurements are level teaspoons of white wheat flour, a teaspoon of either a supplied yeast-water mix or plain water, and either no table sugar (sucrose) or only a small amount. Use as little sugar as the spatula will carry, because a lot of sugar can actually inhibit the yeast.
Rank the treatments from 1 (rises most) to 4 (rises least). If you can, also measure each ball’s height (or diameter) at the start and after incubation, in millimeters; the page will then work out each ball’s percent change and rank the balls by it. Otherwise leave the measurements blank and enter observed ranks directly. Your entries are kept in this browser (see Your Work, on the Fermented Foods tab) and are not sent anywhere.
| Treatment | Predicted rank | Start (mm) | After (mm) | Change | Observed rank |
|---|---|---|---|---|---|
| 1 | – | ||||
| 2 | – | ||||
| 3 | – | ||||
| 4 | – |
In alcoholic fermentation each glucose yields two molecules of CO2 and two of ethanol. Sucrose, table sugar, is a glucose joined to a fructose, which yeast first splits apart, so each sucrose can yield four CO2. The ideal gas law then says how much space that CO2 would take up. Flour contributes some sugar of its own, and the flour’s own enzymes keep releasing maltose from starch as the dough sits, so the calculator lets you include flour sugar as well. Weighing a “small spatula’s worth” of sugar on a milligram balance is the best way to get the first number.
The volume assumes 1 atmosphere of pressure and treats flour sugars as glucose, and is a ceiling rather than a prediction: some CO2 dissolves in the dough’s water, and some escapes from the surface instead of being trapped. The ethanol is the same number of molecules as the CO2 (46.07 g per mole); in real bread most of it evaporates in the oven.
Once your results are in, watch my short conclusion video. The time-lapse video shows what the same process looks like over hours, with a full-sized dough.
In the sense that biologists use the word, fermentation is the use of something other than an electron transport chain to regenerate NAD⁺ for glycolysis. The electrons that glycolysis strips from sugar are handed back to pyruvate, or to another organic molecule made from it, and the result is a fermentation product such as lactic acid or ethanol.
That definition makes sense once you see the problem it solves. Glycolysis, the pathway that splits one glucose into two pyruvates, makes a net two ATP, but it also converts two NAD⁺ into two NADH. A cell contains only a small amount of NAD⁺, so unless the NADH is turned back into NAD⁺, glycolysis soon grinds to a halt. In respiration the NADH passes its electrons to an electron transport chain, which ends with oxygen or some other external electron acceptor. A cell that has no electron transport chain, or no acceptor for one, instead passes the electrons to an organic molecule of its own making, and throws that molecule away.
Two consequences follow. First, a fermenting cell makes its ATP by substrate-level phosphorylation, transferring phosphate directly from an intermediate of glycolysis to ADP, rather than by the electron-transport-driven oxidative phosphorylation of respiration. Second, because the electrons never leave the carbon compounds, fermentation products still contain most of the energy of the original sugar. Ethanol burns, and your liver can use it as fuel; both are reminders of how much energy a fermenting yeast leaves behind (see Why Ferment?).
| Fermentation | Anaerobic respiration | Aerobic respiration | |
|---|---|---|---|
| Final electron acceptor | An organic molecule made from the food itself (pyruvate, acetaldehyde, and so on) | An inorganic molecule other than oxygen, such as nitrate, sulfate, or iron(III) | Oxygen |
| Electron transport chain? | No | Yes | Yes |
| How most ATP is made | Substrate-level phosphorylation | Oxidative phosphorylation | Oxidative phosphorylation |
| ATP per glucose | About 2 (1 to 3, depending on pathway) | Varies with the acceptor; less than aerobic | About 30 (older textbooks say 36 to 38) |
| Examples | Yeast in bread dough; lactic acid bacteria in yogurt; your muscles during a sprint | Denitrifying bacteria in waterlogged soil; sulfate reducers in marine sediments | Your cells most of the time; yeast when sugar is scarce and oxygen present |
Much of the confusion about fermentation comes from the word being used in at least four ways, all of them legitimate in their own settings.
In the 1830s, Charles Cagniard de la Tour, Theodor Schwann, and Friedrich Kützing independently concluded that yeast is a living organism responsible for alcoholic fermentation. The leading chemists of the day, among them Justus von Liebig, disagreed, regarding fermentation as a purely chemical decomposition. Over the following decades Pasteur showed that particular fermentations are the work of particular living microbes, lactic fermentation among them, and that some microbes live without oxygen.
The argument was settled, in a way that made both sides partly right, when Eduard Buchner showed in 1897 that juice pressed from yeast, with no living cells in it, could still ferment sugar. Fermentation was chemistry after all, but the chemistry of enzymes that only living cells make. Buchner received the 1907 Nobel Prize in Chemistry. Arthur Harden and William John Young then found that yeast juice needs phosphate and a small, heat-stable “co-ferment” to work; that co-ferment was later shown, in large part by Hans von Euler-Chelpin, to be NAD, the molecule at the center of this page. Harden and von Euler-Chelpin shared the 1929 Nobel Prize in Chemistry. The full sequence of glycolysis was worked out in the 1930s, largely by Gustav Embden, Otto Meyerhof, and Jakub Parnas, which is why glycolysis is also called the Embden-Meyerhof-Parnas pathway.
When your muscles work harder than your blood can supply them with oxygen, they reduce pyruvate to lactate, exactly as lactic acid bacteria do, and your red blood cells, which have no mitochondria, make all of their ATP this way all the time. The lactate is not wasted: the liver converts it back to glucose, a loop called the Cori cycle after Carl and Gerty Cori. (Lactic acid is also not, as is often said, the cause of the muscle soreness that comes a day or two after hard exercise.)
Measured by ATP per glucose, fermentation is a poor bargain: about 2 ATP, against about 30 for aerobic respiration. So why do so many organisms do it, including some that could respire instead? There are several answers, some firmly established and some still argued over, and they are presented here roughly in that order.
The immediate answer is the one on the What Fermentation Is tab: without some way of turning NADH back into NAD⁺, glycolysis stops, and with it the cell’s ATP supply. Fermentation is the simplest way of doing that, requiring as little as one extra enzyme (lactate dehydrogenase, say) beyond glycolysis itself.
Many places have little or no oxygen: the inside of a dough ball, a wheel of cheese, a sausage, a crock of cabbage under brine, the human large intestine, the mud at the bottom of a pond. An organism living there must either respire with some other electron acceptor or ferment. Some organisms have no choice wherever they live. Most lactic acid bacteria cannot make a complete electron transport chain (some, such as Lactococcus lactis, can respire if they are supplied with heme, which they cannot make for themselves), many clostridia are killed by oxygen, and, as noted, your own red blood cells have no mitochondria.
How much of glucose’s energy does fermentation actually capture? The standard free energy changes below compare the two most familiar fermentations with complete oxidation of glucose.
| Process | ΔG°′ (kJ per mole glucose) | Share of complete oxidation | ATP made |
|---|---|---|---|
| Glucose → 2 lactic acid | −196 | about 7% | 2 |
| Glucose → 2 ethanol + 2 CO2 | −235 | about 8% | 2 |
| Glucose + 6 O2 → 6 CO2 + 6 H2O | −2,840 | 100% | about 30 |
More than 90 percent of the energy thus stays in the products. That is bad news for the fermenting cell but good news for whatever uses the products next: acetic acid bacteria living on the ethanol of fermented fruit, propionibacteria living on the lactic acid left in cheese, your liver recycling lactate, or people drinking beer.
The puzzle deepens because many cells ferment even when oxygen is plentiful. Baker’s yeast given plenty of sugar ferments it to ethanol whether or not oxygen is present, a behavior called the Crabtree effect (after Herbert Crabtree, who in 1929 described sugar suppressing respiration in tumor tissue). Rapidly dividing animal cells, including many cancer cells, similarly turn glucose into lactate despite having oxygen, the Warburg effect, named for Otto Warburg (see Vander Heiden et al., 2009). And fast-growing E. coli excretes acetate even when well aerated, which is known as overflow metabolism.
The most widely discussed explanation is a tradeoff between rate and yield. Fermentation extracts little energy per glucose, but it can extract that energy quickly and with a small investment in enzymes. Basan et al. (2015) argued that, for E. coli, a unit of protein devoted to fermentation produces ATP faster than the same amount devoted to respiration, so that a fast-growing cell short of protein does better, in part, by fermenting. Pfeiffer, Schuster, and Bonhoeffer (2001) pointed out the ecological side of the same tradeoff: when many cells compete for a shared supply of sugar, the cell that uses sugar fast but wastefully can outgrow the cell that uses it slowly but efficiently, even though the efficient strategy would leave everyone better off. They suggested that efficient, high-yield metabolism is favored mainly where cells grow in clusters of relatives that share its benefits.
A second idea is that fermentation products are not just waste but weapons. Ethanol and acids are toxic to many microbes, and the organisms that make them tend to tolerate them better than their competitors do. For Saccharomyces yeasts this is known as the make-accumulate-consume strategy: in a sugary fruit, make ethanol quickly, let it accumulate to levels that inhibit competing microbes, and later, once the sugar is gone, respire the ethanol itself. Thomson et al. (2005) reconstructed the ancestral alcohol dehydrogenase from which yeast’s two modern enzymes arose, one specialized for making ethanol and one for consuming it, and found that the ancestor was suited to making ethanol, consistent with the consuming half of the strategy having evolved later. How much of yeast’s fermenting is explained by warfare rather than by rate remains debated (see Piškur et al., 2006).
Lactic acid bacteria can be viewed the same way. As they ferment, pH falls to levels that most competing organisms cannot tolerate. The succession in fermenting cabbage, in which less acid-tolerant species are replaced by more acid-tolerant ones until the pH is too low for almost anything else, can be seen as this process playing out in a jar.
Food fermentation takes advantage of these microbial strategies. By salting cabbage, which draws out water by osmosis and discourages many competitors, or by keeping milk warm, people arrange conditions so that the organisms whose products we like win, and their acids and alcohol then protect the food from the organisms we do not want. Our relationship with fermentation may also be older than cooking. Carrigan et al. (2015) reconstructed ancestral versions of a primate alcohol dehydrogenase and concluded that about ten million years ago our ancestors’ version became far better at breaking down ethanol, around the time those ancestors began spending more time on the ground, where fallen, fermenting fruit is more common.
Glycolysis, or something close to it, is found in nearly all living things, and the atmosphere contained little oxygen until the Great Oxidation Event some 2.4 billion years ago. An old idea, associated with Alexander Oparin and J.B.S. Haldane, held that the first cells lived by fermenting organic molecules already present in their surroundings. That view is now contested, and many researchers instead favor early cells that made their own organic molecules from carbon dioxide and hydrogen. What seems safe to say is that fermentation is ancient and that it flourished long before there was oxygen to respire.
Fermentations are named for their main products, and most of them differ only in what happens after pyruvate: which molecule ends up accepting NADH’s electrons. The diagram shows pyruvate as a crossroads. Click (or tab to) any of the fermentations to go to its description below; the gray path at the bottom, respiration, is included for contrast. Heterolactic fermentation takes a different route from glucose and so is shown branching off earlier.
Real organisms are rarely confined to one pathway, and the proportions of their products shift with conditions, such as the availability of sugar, pH, and whether any oxygen is present. The pathways are best thought of as the main routes available rather than as fixed recipes.
Lactate dehydrogenase hands the electrons carried by NADH straight back to pyruvate, making lactic acid (lactate). Nothing else is produced, which is what homolactic (“same lactic”) means.
This is the pathway behind most of the sourness in yogurt, many cheeses, fermented sausages, and the later stages of sauerkraut. Lactic acid lowers pH, and falling pH is much of what keeps spoilage organisms and pathogens from growing.
Glucose is not split by the usual glycolytic route. Instead it is converted to a five-carbon sugar phosphate that the enzyme phosphoketolase splits into a three-carbon piece, which becomes lactic acid, and a two-carbon piece, which becomes ethanol or, when another electron acceptor is available, acetic acid. Hetero (“different”) refers to the mixture of products.
The CO2 gives early sauerkraut and kimchi their fizz, and the acetic acid adds to the tang of sourdough. Many heterolactic bacteria can also use fructose as an electron acceptor, reducing it to mannitol.
Pyruvate decarboxylase removes CO2 from pyruvate, giving acetaldehyde, and alcohol dehydrogenase then passes NADH’s electrons to acetaldehyde, giving ethanol.
The CO2 raises bread dough and carbonates beer and sparkling wine; the ethanol is the point of beer, wine, and spirits and the starting material for vinegar. Yeasts also make a little glycerol, which helps balance their electron budget.
Pyruvate is split several ways at once. Some is reduced to lactic acid, some is cut into acetyl-CoA and formic acid, and the acetyl-CoA becomes either ethanol (which consumes electrons) or acetic acid (which yields an extra ATP). The proportions shift to keep electrons balanced.
Mixed acid fermentation matters more in the microbiology laboratory than in the kitchen. Its acids drive the pH low enough to turn the dye methyl red red, which is the basis of the methyl red test used to tell bacteria apart. Relatives of E. coli can be active in the first day or so of some vegetable fermentations before lactic acid bacteria take over.
Two pyruvates are joined and decarboxylated to acetoin, which is then reduced to 2,3-butanediol. Because this route makes fewer acids than mixed acid fermentation, the medium stays closer to neutral.
The Voges-Proskauer test detects acetoin and so identifies bacteria using this pathway. Acetoin and its oxidized relative diacetyl are also made from citrate by some lactic acid bacteria, which is where the buttery flavor of butter, sour cream, and cultured buttermilk comes from.
Propionibacteria can ferment sugars but are known for fermenting the lactic acid that other bacteria leave behind. The propionic acid comes from a cycle that runs through succinate and methylmalonyl-CoA (the Wood-Werkman cycle).
In Swiss-type cheeses such as Emmental, the CO2 collects as the holes (“eyes”) and propionic acid contributes the nutty, sweetish flavor. Propionic acid, as calcium or sodium propionate, is also added to bread to slow mold growth.
Two acetyl-CoA molecules are joined into acetoacetyl-CoA and reduced to butyryl-CoA, which becomes butyric acid. Electrons are also dumped as hydrogen gas. When acids build up, some species switch to making the solvents butanol and acetone instead.
Butyric acid smells of rancid butter and vomit, so in food this pathway usually means spoilage; C. tyrobutyricum growing in hard cheeses produces gas and off-flavors, a defect known as late blowing. The acetone-butanol fermentation was developed by Chaim Weizmann and used to make acetone for British munitions during the First World War.
The processes below are all called fermentations in food science and industry, and all matter for the foods on this page, but none of them regenerates NAD⁺ the way the fermentations above do.
Malic acid, which has two acid groups, is decarboxylated to lactic acid, which has one. Strictly this is a decarboxylation rather than a fermentation, since no electrons are transferred, but winemakers have always called it a fermentation.
Swapping malic acid for lactic acid makes wine taste softer and less tart. Diacetyl made along the way gives some wines, such as many chardonnays, a buttery note.
Acetic acid bacteria oxidize ethanol in two steps, to acetaldehyde and then to acetic acid, passing the electrons to oxygen through an electron transport chain. Because oxygen is the electron acceptor, this is an incomplete aerobic respiration rather than a fermentation, although vinegar makers and food scientists call it acetic acid fermentation.
This is how vinegar is made, and acetic acid bacteria also contribute to kombucha, kefir, and cacao. A true acetic acid fermentation does exist, made by anaerobic bacteria such as Moorella thermoacetica that turn one glucose into three acetic acids, but it plays no part in food.
Several traditional food processes are called fermentations because microbes transform the food, even though no fermentation in the biochemical sense is involved. Molds such as Aspergillus (koji) and Rhizopus (tempeh) and the bacterium Bacillus subtilis (natto) grow aerobically, respiring as they digest proteins and starches with the enzymes they release.
These processes still count as fermented foods under the food definition described on the What Fermentation Is tab, which is about microbes transforming food rather than about how the microbes make their ATP.
In black tea, crushing the leaves lets the leaf’s own enzyme, polyphenol oxidase, oxidize its polyphenols with oxygen from the air; no microbes are involved at all. In fish sauce and its Roman ancestor garum, enzymes from the fish’s own guts and flesh digest the fish over many months in heavy salt, a self-digestion called autolysis. Salt-tolerant bacteria contribute acids, aroma, and some enzymes of their own, but the salt keeps most microbes from growing, and the fish’s enzymes do most of the digesting.
Neither is a fermentation in the biochemical sense, and black tea is not even a fermented food by the food definition, since no microbes are involved. Fish sauce is more ambiguous: it is conventionally counted among fermented foods, but most of its transformation is the work of the fish’s enzymes rather than of microbes. Enzymes matter in many fermented foods; what sets these two apart is that the enzymes are the food’s own.
Every fermentation product is, in effect, a place where a cell has parked the electrons it removed from sugar. The table lists the main ones, the pathways that make them, the foods on the Fermented Foods tab in which they matter, and what they do there. Most organic acids are listed under their acid names; inside cells, at near-neutral pH, they exist mainly as the ionized forms given in parentheses (lactate, acetate, and so on).
| Product | What it does, what makes it, and where it is found |
|---|---|
| Carbon dioxide CO2 |
Leavens bread, carbonates beer, sparkling wine, kefir, and kombucha, and forms the holes in Swiss cheese. In alcoholic fermentation it is released when pyruvate is decarboxylated to acetaldehyde. |
| Ethanol C2H5OH |
The alcohol of beer, wine, and spirits, and the starting material that acetic acid bacteria oxidize to vinegar. In bread dough it mostly evaporates during baking. |
| Lactic acid (lactate) CH3CH(OH)COOH |
The main acid of yogurt, sour cream, cheese, sauerkraut, kimchi, fermented pickles, and fermented sausages. By lowering pH it curdles milk proteins and inhibits many spoilage organisms and pathogens. Your muscles make it too. |
| Acetic acid (acetate) CH3COOH |
The acid of vinegar, and a contributor to the tang of sourdough, kimchi, and kombucha. Most acetic acid in food comes from acetic acid bacteria oxidizing ethanol, which is not strictly a fermentation. |
| Propionic acid (propionate) CH3CH2COOH |
Gives Swiss-type cheeses part of their nutty, sweetish flavor. As calcium or sodium propionate it is added to bread as a mold inhibitor.
Made by: Propionic
Found in: Cheese |
| Butyric acid (butyrate) CH3CH2CH2COOH |
Smells of rancid butter; in food, butyric acid fermentation is usually a sign of spoilage. Butyrate is also made by bacteria in the human large intestine, where it serves as fuel for the cells lining the colon.
Made by: Butyric
Found in: – |
| Butanol C4H9OH |
A solvent, and a candidate biofuel, made by Clostridium acetobutylicum and relatives in the acetone-butanol fermentation.
Made by: Butyric
Found in: – |
| Acetone (CH3)2CO |
The solvent of nail polish remover. Its fermentative production, developed by Chaim Weizmann, was industrially important during the First World War.
Made by: Butyric
Found in: – |
| Formic acid (formate) HCOOH |
Made when pyruvate is cut into acetyl-CoA and formate. Under acid conditions E. coli splits it into hydrogen and carbon dioxide, the source of the gas bubbles seen in some laboratory tests.
Made by: Mixed acid
Found in: – |
| Succinic acid (succinate) HOOC(CH2)2COOH |
A minor product of mixed acid fermentation, and an intermediate on the route by which propionibacteria make propionic acid.
Made by: Mixed acid
Found in: – |
| Hydrogen H2 |
A way of getting rid of excess electrons without making another organic product. Hydrogen from clostridia contributes to the gas of late-blown cheeses.
Made by: Mixed acid Butyric
Found in: – |
| 2,3-Butanediol CH3CH(OH)CH(OH)CH3 |
The main product of butanediol fermentation; present in small amounts in wine. Of industrial interest as a starting material for plastics and fuels.
Made by: Butanediol
Found in: – |
| Acetoin and diacetyl CH3COCH(OH)CH3; CH3COCOCH3 |
Buttery-tasting compounds. Some lactic acid bacteria make them from citrate, giving butter, sour cream, cultured buttermilk, and some wines their buttery note; in beer, diacetyl is considered an off-flavor. Acetoin is what the Voges-Proskauer test detects.
Made by: Butanediol Malolactic
Found in: Wine Sour cream and cultured buttermilk |
| Acetaldehyde CH3CHO |
The intermediate between pyruvate and ethanol in alcoholic fermentation, and the compound largely responsible for the characteristic “green apple” aroma of plain yogurt, where it is made by the yogurt bacteria.
Made by: Alcoholic
Found in: Yogurt |
| Glycerol C3H5(OH)3 |
A minor product of yeast fermentation, made to reoxidize NADH that is not balanced by ethanol production. It contributes to the body of wine.
Made by: Alcoholic
Found in: Wine |
| Mannitol C6H8(OH)6 |
A sugar alcohol made when heterolactic bacteria such as Leuconostoc use fructose as an electron acceptor. It contributes a mild sweetness to fermented vegetables such as sauerkraut and kimchi.
Made by: Heterolactic
Found in: Sauerkraut Kimchi |
| Gluconic acid HOCH2(CHOH)4COOH |
Made when acetic acid bacteria oxidize glucose, again not strictly by fermentation. One of the acids of kombucha.
Made by: Acetic (oxidation)
Found in: Kombucha |